Rotating Wing Mast – theoretical discussion

Discussion in 'Sailboats' started by Man Overboard, Nov 15, 2006.

  1. rxcomposite
    Joined: Jan 2005
    Posts: 2,752
    Likes: 608, Points: 113, Legacy Rep: 1110
    Location: Philippines

    rxcomposite Senior Member

    Petereng,
    Thank you for thoughts on the alternative design. You, as a proven mast designer and composite engineer, have a deeper knowledge and understanding on the topic and your input is highly appreciated. It was suggested when the discussion moved towards slimmer design for better aerodynamic performance against the stiffer, more rounded section. That is why I am a latecomer. I was following your inputs in the latter part and your posts are highly informative.

    True, the top skin takes the most stress as it is furthest away from the neutral axis. In this case, as I have mentioned, the spar takes the bending and shear load with the skin acting only in torsional load. In span buckling is controlled by ribs. I did not mention this as this is a preliminary discussion.

    This type of construction is proven, tested to destruction, and FEA analyzed (In fact, I was going to suggest to Erwan if you can do preliminary FEA). As shown in the images, the panels, top skin, core, and bottom skin was FEA analyzed. Although the preliminary sizing technique illustrated in the book was very simple and conservative as the author admits, it lends well to a more sophisticated quadratic analysis as this is basically a wide flange H beam.

    It is in production and one of the licensed manufacturer is in Australia. Later model used carbon fiber skin to close the strain gap between the Eglass and CF resulting in a lighter structure.

    It is a proven method in wing design as thousands are already flying using this method. It was a precursor to the successful WingSail by Cogito, an AC contender as featured in PB issue 39. The skin is mylar, the ribs are hollow and much closer together. The root design of spar and skin remains the same.

    Illustrations by M. Hollmann
     

    Attached Files:

    Last edited: Oct 10, 2013
  2. rxcomposite
    Joined: Jan 2005
    Posts: 2,752
    Likes: 608, Points: 113, Legacy Rep: 1110
    Location: Philippines

    rxcomposite Senior Member

    I think this was covered in E. Sponberg numerous articles.

    If it is a simple circular section with varying diameter and thickness to maintain stiffness relative to load, it will be relatively simple. You said in one of your post it is conical. The only complicated part is the Euler's principles on columns (one part stayed, the other part unstayed) and the elastic load of the stay due to;

    a. the elastic property of the cable
    b. the longitudinal and lateral bending of the hull and deck due to tension of the cable. Picture this as a bow bending when the string is pulled back.
     
  3. petereng
    Joined: Jan 2008
    Posts: 581
    Likes: 22, Points: 28, Legacy Rep: 252
    Location: Gold Coast Australia

    petereng Senior Member

    RX - We have now moved to "Wing" design and this is a new philosophical arena compared to small roundish conventional masts. Yes indeed aircraft have used shear webs and I beam construction for many years. The I beam takes the bending loads and the skin resists the torsional loads (I beams are poor in torsion but this is whats needed in a sailing wing). Plus it gives the structure a bit of redundency in case local failures occur. All well and good, a relatively highly developed and understood solution. Personally I feel the current status quo of boat "wing" design can be improved considerably. If you look at a hang glider wing it has no internal ribs at all, only shaped battens. The aero loads are very similiar to sailing loads and the areal weights of them are similiar as well. The on-edge honeycomb or foam ribs are convenient to make on CNC routers so are commonly used, but are very delicate. I believe it would be much better to make moulded battens. These would be more reliable then the current type of rib. The biggest problem in designing wings is getting them to twist properly. If you go to my website I have various structural animations of wings twisting. Happy to chat about any structural stuff. Groper yopu will note in the wing article that several of the tubes were made on male aluminium mandrels... Peter S
     
  4. rxcomposite
    Joined: Jan 2005
    Posts: 2,752
    Likes: 608, Points: 113, Legacy Rep: 1110
    Location: Philippines

    rxcomposite Senior Member

    Yes Petereng, I realized late that I have moved to a more sophisticated design.

    I have read battens and kevlar inserted in this new and sophisticated laminated sail.

    I will take a look in your website. Should be interesting. My research took me to design of wing blades as installed on wind generators. It is about building in twists and "droop" into the blades so that when it reaches its optimum RPM, it aligns to the optimum AoA. Similar to the mast but with the absence of a centrifugal force that tends to straighten the blade.
     
  5. petereng
    Joined: Jan 2008
    Posts: 581
    Likes: 22, Points: 28, Legacy Rep: 252
    Location: Gold Coast Australia

    petereng Senior Member

    Rx - in regard to wind turbine baldes. Now they are at the length limit of "solid" technology the manufactures are looking at soft skinned versions just like the boat wings. Things go in circles, right back to how the dutch did it!! You are speaking about aeroelastic coupling, the "droop" you speak of is usually incorporated so that as the blades load up they do not hit the pylon!! Opps I've seen some development turbines where the balde is on the rear side of the pylon so this effect can be neglected. Makes the tooling much easy to build. Drooped turbine blade moulds are huge!! Peter
     
  6. rxcomposite
    Joined: Jan 2005
    Posts: 2,752
    Likes: 608, Points: 113, Legacy Rep: 1110
    Location: Philippines

    rxcomposite Senior Member

    That would be catastrophic if they designed the droop taking into consideration only the uppermost wingblade in the vertical postion.:p:D

    Which leads to another question I have always wanted to ask about. They say the wind is faster as you go higher (up the mast only) as the wind in sea level is affected by surface turbulence. Is this correct?
     
  7. petereng
    Joined: Jan 2008
    Posts: 581
    Likes: 22, Points: 28, Legacy Rep: 252
    Location: Gold Coast Australia

    petereng Senior Member

    RX _ At land or sea level the wind speed is near zero. At abou8t 30m high the wind speed is about at its free velocity. This is called wind shear. This is because the land and water rub against the air and this friction slows the air down at sea or land level. This is documented in various wind design codes and sailing books. This is one reason we need to twist sails and aircraft do not. As the apparent wind velocity is different as the height increases the sail has to be twisted to get best efficiency. This is the same as propellors and wind turbine blades. As they are rotating about an axis the apparent wind velocity is proportional to the radius. So the blade has to be twisted to have the correct AoA along the blade. Cheers Peter s
     
  8. rxcomposite
    Joined: Jan 2005
    Posts: 2,752
    Likes: 608, Points: 113, Legacy Rep: 1110
    Location: Philippines

    rxcomposite Senior Member

    Thanks Peter.:cool:
     
  9. groper
    Joined: Jun 2011
    Posts: 2,483
    Likes: 144, Points: 73, Legacy Rep: 693
    Location: australia

    groper Senior Member

    Peter, i wish you join in my thread about hydrofoil assist... i could use some better theoretical help to determine the feasability of foil assist on my catamaran under construction! Im sure many things are possible if enough effort is put into solving the problems of construction... Most are quick to say its impossible or it wont work, and maybe it fails for many people but that often means they simply didnt solve all the problems... Im not even sure what we are theorizing about in this thread anymore??? :)
     
  10. Erwan
    Joined: Oct 2005
    Posts: 460
    Likes: 28, Points: 28, Legacy Rep: 97
    Location: France

    Erwan Senior Member

    Thank you all for your contribution , as you are going too fast for my old brain to keep up, I ll continue to follow with no much things to post.

    Thanks Peter for the better definition of wingsail, I should have done before in order to avoid to confuse RX.

    Regarding the wind gradient, you'll find attached a workpaper, you can find relevant keywords in order to googlize for more.

    Groper , I ll vist your thread about foil control, here is a link with Melvin: Morelli interview who explained why AC72 achieved this foil configuration, basically because for the writting of the rule box, Italian refuse "moveable parts" on the foils for cost reason, as aresult they achieve moveable foils+ foil casing.

    He mention some research using car-shock absorber to address stability with moveable part concepts.

    http://www.cupinfo.com/en/americas-cup-gino-morrelli-foils-multihulls-13144.php

    So mimicking blindly AC system is probably not the only direction to explore

    Cheers Everybody

    Erwan
     

    Attached Files:

  11. rxcomposite
    Joined: Jan 2005
    Posts: 2,752
    Likes: 608, Points: 113, Legacy Rep: 1110
    Location: Philippines

    rxcomposite Senior Member

    Thanks for the link Erwan. That was the missing link.

    I think everybofy got confused following a long thread.

    The first part of the thread talked about round rotating sail mast and its inefficiency to air flow.

    Immediately following, Eric gave the option of a conical mast with a spar. You said "conical" so I posted in Excel image a conical method of construction using the spar method but still a sail mast. It is a thin conical mast that will support the sail. In effect a thin highly curved airfoil as illustrated in the Excel image.

    Now a thin highly curved airfoil has the same lift as a thick high lift, high drag airfoil. With a thicker section (spar height), the mast/spar can be stiffer, and is more efficient leading to the design of wing sail. a more advanced method than originally discussed in the first and second part.;)
     
  12. Eric Sponberg
    Joined: Dec 2001
    Posts: 2,021
    Likes: 248, Points: 73, Legacy Rep: 2917
    Location: On board Corroboree

    Eric Sponberg Senior Member

    I agree with Peter that in free-standing wingmast design, cored skins do not make sense for the very reason of skin buckling that he states. Also, for whatever skin thickness you need to handle all of the bending loads, adding a core to the skin only adds weight, cost, and building complication (more labor hours) and it does not reduce skin thickness at all. In fact, it may increase skin thickness, therefore more material, more weight, and more cost. And I agree that if you want to make the lightest possible mast, then it should be bigger rather than smaller to keep the skins as light as possible. However, making the mast bigger in section also creates more drag when sailing, and more windage when at anchor.

    Therefore, you want the smallest aerofoil section that is consistent with strength and tip deflection. If you go too small, the mast gets too heavy (thicker wall) and the tip deflection gets too great and the sail inverts easily. These are soft sail wingmasts I am talking about, where the sail can be taken down when the boat is put to bed.

    The governing factor in all of this, as Peter mentions, is what I use as thickness-to-diameter ratio, the inverse his radius-to-thickness ratio. t/D is easier for me to understand and is the factor I use in my wingmast designs. Peter's radius to thickness ratio = 50 translates to a diameter to thickness = 100. Invert that, and you have a t/D = 0.01. That works for metal structures because metals are isotropic--they have the same mechanical properties in all directions. Composites don't have that--their mechanical properties vary in different directions. I have found in my experience, based on a lot of full-scale breaking of composite tubes and analyzing the breaking and buckling behavior, that the minimum acceptable t/D ratio is three times the above ratio for metals to make up for the orthotropicity of composites. That is, t/D => 0.03. This applies to laminates that are at least 60/40 UDR to off-axis fiber. For laminates that are more 50/50 UDR/off-axis, they are more isotropic and would tend back toward t/D = 0.01. But you lose the overall strength in the laminate because not enough fiber is running axially to handle the mast bending load, and too much fiber is running circumferentially where you don't need that much of it. This is why in sailboat free-standing masts, I have gone to minimum 60/40 split on fiber orientation for round section masts, and max out at 80/20 split for wing-shaped masts. By the way, these ratios apply to wingmasts in the short direction, that is, across the width of the section, not across the length, and that is because the width direction is the weaker bending direction for the wing. In real life, the masts bend some back, some to the side.

    Finally, with regard to shear webs. I always put a shear web in my wingmast designs to handle the shear between the tension and compression sides of the mast. The more round the mast section is, the less need for a shear web. This shear web is easy to make with a relatively thin core which does add to its stability inside, and so is the one place where I do use cores in my wingmast designs. The skins on the web are fully ±45° orientations because they are taking pure shear between the mast side walls, and therefore those fibers run in exactly the right direction to handle the shear loads.

    Eric
     
    1 person likes this.
  13. petereng
    Joined: Jan 2008
    Posts: 581
    Likes: 22, Points: 28, Legacy Rep: 252
    Location: Gold Coast Australia

    petereng Senior Member

    Hi Eric - So you have caught up with #161? Happy to pick up where we left off and do a mast panel comparison if you like. What we are up to now I think is I have shown how the stack is important now we move onto geometry considerations. ie the flatter or more oval the shape the more it has a tendency to hinge or cripple. The solution with this is a shear web or making it thicker. 3 or 4:1 conventional sections IMO don't need webs but at some point a wing section changes from a geometric self supporting section to a shape that needs to be subdivided or internally ribbed or both eg a plane wing. Sigurd I'll try to digest your geometry this week as well. Been busy trying/selling one of my skiffs over the last few weeks. I've now sold it, so have a space for a new project. Cheers Peter S
     
    1 person likes this.
  14. rxcomposite
    Joined: Jan 2005
    Posts: 2,752
    Likes: 608, Points: 113, Legacy Rep: 1110
    Location: Philippines

    rxcomposite Senior Member

    Took some coaxing for Eric to come back.

    In preliminary mast sizing techniques, the rule of mixture gives good results in predicting stiffness. That is, when there are two or more laminates running in different ply directions. A laminate is stiffest when the fibers are aligned in the direction of stress and gradually loses its stiffness or modulus when rotated from its natural fiber direction.

    Take the case of a round mast that is subjected to bending and torsional stress. The axial layers of unidirectional fibers are strongest at 0 degrees and the +45-45 biax strongest at 45 degrees. Referring to the Engineering Constants graph, the carbon Uni is best at 0 degree load and the Eglass biax at 45 degree load. The uni’s are sized to handle the bending load and the biax to take care of the torsional load with the corresponding number of layers (thickness) to satisfy the load requirement. The Uni/biax ratio, as Eric explained is what has been found to work best.

    In filament winding, where the helical angle has more freedom of control, the optimum helical angle can be derived by the torsion formula if the load and the lever arm are known.

    In theory, the outermost fibers away from the neutral axis receive the most stress and intuition tells us that the uni’s should be in the outermost layer to function most efficiently. This is correct but in practical application, it is the outermost layer that flexes considerably more and needs to be of low modulus to prevent fatigue. It also forms a skin where it holds together the unidirectional fibers to prevent it from unraveling. Notice that if carbon fiber is chosen, it is most sensitive to off axis loads and degradation starts at 7 degrees. If the axial fibers were of carbon, it will carry or limit the bending load. The outermost helical fibers, being of low modulus will just bend and be limited from flexing too much by the carbon uni. If the unidirectional fibers were of low modulus and not sized properly, it would bend more and transfer the stress to the outermost fiber causing premature failure.

    In designs where a spar is utilized, carbon fiber is used in the spar caps to control the bending and Eglass biax to handle the web shear and torsional load (see illustration).

    This can easily be demonstrated by loading a cantilevered H beam on one end. Glue a piece of plastic on one face and apply load until it beam deflects. The low modulus plastic will just follow the bend. If a thin sheet of solid glass is glued instead, the glass will shatter as the beam bends.

    There has been concern over the use of cores due to buckling. Note that with large diameter mast, a core as low as 2 lbs/ft2 can be used. This weighs almost nothing and only serves as form. Cored mast panels/sides are high aspect ratio and high aspect ratio panels flex. It is impractical to use a higher density core as this will make the overall construction heavy. Along the span, webs, ribs, or bulkheads can be added shortening the span of the panel. The web also acts as local reinforcement in way of coachroof support or boom connection.

    Core is a quick way to stiffen the structure but becomes impractical when the cross sectional area becomes more ovalish. With the long sides coming closer towards the neutral axis, a heavier laminate is needed for the face and a higher density core is required.
     

    Attached Files:

    1 person likes this.

  15. petereng
    Joined: Jan 2008
    Posts: 581
    Likes: 22, Points: 28, Legacy Rep: 252
    Location: Gold Coast Australia

    petereng Senior Member

    Incorrect - if the structure is a cored thin skins and thick low density cores are the correct answer not thick skins. Again please use correct sandwich and laminate theory to determine answers. Diab have an excellent technical manual on sandwich design. There is a difference in theory between thin skin sandwiches and thick skin be careful. You must ensure the skin rigidity adequate to ensure the core is only in shear. If you use a thin core and thick skins you can fail the core easily. Loo upo the diab publication Cheers Peter S
     
Loading...
Forum posts represent the experience, opinion, and view of individual users. Boat Design Net does not necessarily endorse nor share the view of each individual post.
When making potentially dangerous or financial decisions, always employ and consult appropriate professionals. Your circumstances or experience may be different.